We uniformly analyze 136 optically detected PNe and candidates from the GLIMPSE-I survey in order to to develop robust, multi-wavelength, classification criteria to augment existing diagnostics and provide pure PN samples. PNe represent powerful astrophysical probes. They are important dynamical tracers, key sources of ISM chemical enrichment, windows into late stellar evolution, and potent cosmological yardsticks. But their utility depends on separating them unequivocally from the many nebular mimics which can strongly resemble bona fide PNe in traditional optical images and spectra. We merge new PNe from the carefully evaluated, homogeneous MASH-I and MASH-II surveys, which offer a wider evolutionary range of PNe than hitherto available, with previously known PNe classified by SIMBAD. Mid-infrared (MIR) measurements vitally complement optical data because they reveal other physical processes and morphologies via fine-structure lines, molecular bands and dust. MIR colour-colour planes, optical emission line ratios and radio fluxes show the unambiguous classification of PNe to be complex, requiring all available evidence. Statistical trends provide predictive value and we offer quantitative MIR criteria to determine whether an emission nebula is most likely to be a PN or one of the frequent contaminants such as compact HII regions or symbiotic systems. Prerequisites have been optical images and spectra but MIR morphology, colours, environment and a candidate's MIR/radio flux ratio provide a more rigorous classification. Our ultimate goal is to recognize PNe using only MIR and radio characteristics, enabling us to trawl for PNe effectively even in heavily obscured regions of the Galaxy.